Using Multiple Diacritics in Arabic Scripts for Steganography By Yousef Salem Elarian Aleem Khalid Alvi 1

Contents Introduction The classification tree of steganography Characteristics of the Arabic Script Related Work ApproachesEvaluationConclusion 2

Introduction Stenography is the approach of hiding the very existence of secret messages. 3

The classification tree of steganography classification tree Figure 1: classification tree 4

Characteristics of the Arabic Script The Arabic alphabet has Semitic origins derived from the Aramaic writing system. Arabic diacritic marks decorate consonant letters to specify (short) vowels. Dots and connectivity are two inherent characteristics of Arabic characters. Arabic basic alphabet of 28 letters, 15 have from one to three points, four letters can have a Hamzah, and one, ALEF, can be adorned by the elongation stroke, the Maddah 5

contd.. Little has been proposed for Arabic script steganography. Two inherent properties of Arabic writing: dots and connectability. Some work needs new fonts. Other introduces using the existent Kashidah, helped with dotted letters, instead. Hamzated characters can also be used as along with dotted ones. Figure 2. Arabic diacritic marks 6

The main idea: Diacritics Notice the differences in levels of darkness (as to the right) or colors (as below) in the single and the repeated diacritics. 7

Scenarios 1 st scenario (Secret = ) –Direct (block size = inf.) For each block b i = n d Repeat the i th diacritic n d times. Repeat the 1 st diacritic 49 extra times! –Blocked (block size = 2) Repeat the 1 st diacritic 3 extra times (3 = (11) b ); the 2 nd one, 0 extra times (0 = (00) b ); and the 3 rd one, 1 extra time (1 = (01) b ). 8

Scenarios Cntd RLE While(secret!=EOF & cover!=EOF b = b^ While(secret.b = b) Type diacritic –For Secret = Repeat the 1 st diacritic 2 times (1’s in (11) b ); the 2 nd one, 3 times (0’s in (000) b ); and the 3 rd one, 1 time (for 1). 9

Related Work Shirali-Shahreza: –The position of dots is changed to render robust, yet hidden, information. The method needs special fonts. Gutub: –Secret-bit hiding after dotted letters by inserting Kashidah’s. A small drop in capacity occurs due to restriction of script on Kashidah and due to the extra- Kashidahs. Aabed et al.: – Redundancy in diacritics is used to hide information by omitting some diacritics. 10

The textual approach The textual approach chooses a font that hides extra (or maybe all) diacritic marks completely. It uses any encoding scenario to hide secret bits in an arbitrary number of repeated but invisible diacritics. A softcopy of the file is needed to retrieve the hidden information (by special software or simply by changing the font). 11

contd.. ScenarioApproachExtradiacritics 1 st scenario (stream) Repeat the first diacritic 49 times. (49 = (110001) b ) st scenario block size=2 Repeat the first diacritic 3 times (3 = (11) b ), the second one 0 times (0 = (00) b ), and the third one 1 time (1 = (01) b ) = 4. 2 nd scenario (RLEstart=1) Repeat the first diacritic 2 times (2 = number of 1’s in (11) b ), the second one 3 times (3 = number of 0’s in (000) b ), and the third one 1 time (for 1). (2-1) + (3-1) + (1-1) = 3. Table 1. The encodings of the binary value according to the two scenarios of the first approach 12

The image approach, The image approach selects one of the fonts that slightly darken multiple occurrences of diacritics. This approach needs to convert the document into image form to survive printing. This unfortunate fact reduces the possible number of repetition of a diacritic to the one that can survive a printing and scanning process (up to 4 as the last two columns of the first diacritic 13

Approaches cont.. Figure 4: The degree of brightness of the diacritic marks repeated 1, 2, 3, 4 and 5 times each 14

ApproachCapacityrobustnesssecurity Text + softcopy High, up to infinity in 1 st scenario Not robust to printing Invisible, but in the code Image + softcopy Very low, due to image overhead Robust to printing Slightly visible. Sizeable Image + hardcopy Moderate, 1 st scenario, blocks of 2 Robust to printing Slightly visible Approaches cont.. Table 2. Comparison between the two approaches in terms of capacity, robustness and security. 15

Evaluation Approachpqr(p+r+q)/2 Dots Kashidah-Before Kashidah-After Diacritics Table 3. The ratios of usable characters for hiding both binary levels according to the three studied approaches 16

Comparison Diacritics vs. Kashidah Pros: –While Kashidah suffers from restrictions on its insertion, almost every character can bare a diacritic on it. This disadvantage of the Kashidah method becomes severer for the need of dotted character. Cons: –Diacritics never come alone; but with another character  a stable overhead of 2 bytes per secret-baring position. 17

Comparison Cntd. The Advantage The advantage of our work is that each usable character can bare multiple secret bits with 1 character as overhead. Although this same overhead can be claimed in the Kashidah method, it can’t really be applied for Kashidah becomes too long and noticeable. 18

Conclusion The text and image approaches are discussed which are used to hide information in Arabic diacritics for steganographic. This paper presents a variety of scenarios that may achieve up to arbitrary capacities. Sometimes tradeoffs between capacity, security and robustness imply that a particular scenario should be chosen. The overhead of using diacritics was, experimentally, shown very comparable to related works. The advantage of the method is that such overhead decreases if more than one diacritical secret bit is used at once. 19

The End 20